IJMCM Original Article Autumn 2016, Vol 5, No 4 Bioinformatics Analysis Reveals Genes Involved in the Pathogenesis of Ameloblastoma and Keratocystic Odontogenic Tumor Eliane Macedo Sobrinho Santos1,2, Hércules Otacílio Santos3, Ivoneth dos Santos Dias4, Sérgio Henrique Santos5, Alfredo Maurício Batista de Paula1, John David Feltenberger6, André Luiz Sena Guimarães1, Lucyana Conceição Farias¹ 1. Department of Dentistry, Universidade Estadual de Montes Claros, Minas Gerais, Brazil. 2. Instituto Federal do Norte de Minas Gerais-Campus Araçuaí, Minas Gerais, Brazil. 3. Instituto Federal do Norte de Minas Gerais-Campus Salinas, Minas Gerais, Brazil. 4. Department of Biology, Universidade Estadual de Montes Claros, Minas Gerais, Brazil. 5. Department of Pharmacology, Universidade Federal de Minas Gerais, Brazil. 6. Texas Tech University Health Science Center, Lubbock, TX, USA. Submmited 10 August 2016; Accepted 10 October 2016; Published 6 Decemmber 2016 Pathogenesis of odontogenic tumors is not well known. It is important to identify genetic deregulations and molecular alterations. This study aimed to investigate, through bioinformatic analysis, the possible genes involved in the pathogenesis of ameloblastoma (AM) and keratocystic odontogenic tumor (KCOT). Genes involved in the pathogenesis of AM and KCOT were identified in GeneCards. Gene list was expanded, and the gene interactions network was mapped using the STRING software. “Weighted number of links” (WNL) was calculated to identify “leader genes” (highest WNL). Genes were ranked by K-means method and Kruskal- Wallis test was used (P<0.001). Total interactions score (TIS) was also calculated using all interaction data generated by the STRING database, in order to achieve global connectivity for each gene. The topological and ontological analyses were performed using Cytoscape software and BinGO plugin. Literature review data was used to corroborate the bioinformatics data. CDK1 was identified as leader gene for AM. In KCOT group, results show PCNA and TP53. Both tumors exhibit a power law behavior. Our topological analysis suggested leader genes possibly important in the pathogenesis of AM and KCOT, by clustering coefficient calculated for both odontogenic tumors (0.028 for AM, zero for KCOT). The results obtained in the scatter diagram suggest an important relationship of these genes with the molecular processes involved in AM and KCOT. Ontological analysis for both AM and KCOT demonstrated different mechanisms. Bioinformatics analyzes were confirmed through literature review. These results may suggest the involvement of promising genes for a better understanding of the pathogenesis of AM and KCOT. Downloaded from ijmcmed.org at 17:06 +0330 on Tuesday October 5th 2021 [ DOI: 10.22088/acadpub.BUMS.5.4.199 ] Key words: Ameloblastoma, keratocystic odontogenic tumor, cell proliferation, apoptosis, leader gene Corresponding author: Universidade Estadual de Montes Claros Hospital Universitário Clemente de Faria Laboratório de Pesquisa em Saúde Avenida Cula Mangabeira, 562 Montes Claros, Minas Gerais, Brasil. E-mail: [email protected] Santos EMS et al. dontogenic tumors consist of a heterogeneous gene deregulations in these tumors. Silico Ogroup of lesions that originate from the tissue approaches, such as bioinformatic analysis, have that forms the teeth (1). These tumors affect been performed to investigate signaling pathways, individuals in different age groups, involving protein interactions, microRNA prediction models, mandibular and maxillary region, with central or and gene expression to obtain the best peripheral location. Some lesions are asymptomatic understanding of pathological mechanisms of and are discovered by chance through routine diseases (13). The computational method is an radiographs. Additionally, odontogenic tumors important tool to understand molecular aspects of could promote the local expansion or facial oral pathology and medicine (14-16). swelling (2, 3). Pathogenesis of odontogenic tumors This study aimed to investigate the differential is not well known. Several studies were performed involvement of protein-coding genes in the to identify genetic deregulations and molecular pathogenesis of AM and KCOT, through alterations in an attempt to explain the mechanisms bioinformatics analysis. of oncogenesis, cytodifferentiation, and tumor progression (3, 4). Materials and methods Ameloblastoma (AM) is a benign tumor Bioinformatics and biological systems analysis originating in the odontogenic epithelium without Initially, key genes involved in the ectomesenchyme, affecting the maxillo-mandibular pathogenesis of AM and KCOT were identified by complex (5). It is an asymptomatic lesion, and it searching the GeneCards database (17). The gene presents locally invasive behavior, and higher nomenclature adopted was defined by Human recurrence rates (6). The differential diagnosis Genome Organization (HUGO). The keywords, includes a variety of odontogenic cysts and tumors, chosen according to Medical Subject Headings particularly keratocyst odontogenic tumor and (MeSH), were “ameloblastoma and gene myxoma, non-odontogenic tumors and cysts, as expression” and “keratocystic odontogenic tumor central giant cell lesions and fibro-osseous lesions and gene expression”. (7, 8). After this step, a list of potential “candidate The keratocystic odontogenic tumor (KCOT), genes” related to AM and KCOT was generated to according to the most recent classification of each tumor. Then, this gene list was expanded using tumors of the head and neck of the World Health the web-available software STRING (version 9.1) Organization (WHO), has been categorized as (14), mapping the interaction network between benign neoplasm derived from odontogenic these protein-coding genes. Direct and indirect gene epithelium. The great clinical relevance of KCOT is interactions were considered with a high degree of related to aggressive clinical behavior, high recu- confidence (above 0.9, range 0-0.99) (14). With this rrence and proliferation rate (9, 10). However, there process, new genes linked to AM and KCOT could are still disagreements, questioning whether this be identified. For every gene interaction identified, odontogenic lesion indeed is a neoplasm or a cyst we summed the interaction score of each gene, of odontogenic nature (11). Some studies have generating a combined association score. This score Downloaded from ijmcmed.org at 17:06 +0330 on Tuesday October 5th 2021 [ DOI: 10.22088/acadpub.BUMS.5.4.199 ] sought to understand these aspects through mole- was adjusted, multiplying it by 1,000 (14), to obtain cular investigations (11, 12). a single value called weighted number of links Despite efforts focused on understanding the (WNL). The genes that showed the largest WNL pathogenesis of odontogenic tumors, little is known values were named “leader genes” (14). Total about the real influence of molecular pathways and interaction score (TIS) was also calculated using all Int J Mol Cell Med Autumn 2016; Vol 5 No 4 200 Bioinformatics Analysis: Ameloblastoma and Keratocystic Odontogenic Tumor interaction data generated by the STRING database according to the following inclusion criteria: to achieve global connectivity for each gene articles published in English, with availability in its involved in the process (14). The value of full version, whose study has been conducted with WNL/TIS ratio represents the most influential laboratory experiments and that addressed the topic genes in the network (specificity score). Genes with of interest. After the initial selection of material, no link (orphan genes) were excluded from this items with incompatible content with the object of analysis. the study were excluded. The final research Genes were ranked according to this material was made up of 66 selected papers for AM parameter in clusters, by the clustering method K- and 25 for KCOT. The synthesis of the collected means. The number of clusters was calculated using data and evidence analysis are based on information the following equation: Cluster number =TETO regarding the characteristics, methods and study (LOG(CONT.NÚM(N);2);1). The number of clust- endpoints (title, aim, methodological design, ers was obtained when mathematical convergence sample, results, conclusion and level of evidence). was achieved. To evaluate the differences among various classes based on WNL, Kruskal-Wallis test Results was used. Statistical significance was set at a p- Survey of genes associated with AM and KCOT value <0.001. Interacting genes were classified as pathogenesis and interaction network up-regulated or down-regulated, as previously The search through GeneCards and String described (14), to each type of odontogenic tumor, database included 119 genes related to AM and AM or KCOT. Complementary analysis of 54 genes related to KCOT. Figures 1a and 1b biological systems was performed by topological show gene interaction maps and also increased and ontological analysis. The first was carried out and decreased gene expression in AM and with Cytoscape software (18), and ontological KCOT, respectively. This figure also reveals the analysis was performed with BinGO plugin (14). orphan genes. In the gene interaction network for Literature review AM, genes were the following: AIFM1, AMBN, A literature review was performed according AMELX, AMELY, BCL2L15, CADM1, CALB2, to inclusion and exclusion criteria of the Preferred CD68, CHKA, CLDN4, CLDN5, DSPP, ENAM, Reporting Items for Systematic Reviews and HOTAIR,